Human-in-the-loop QA: let an AI agent run your test sheet on Android, iPhone, web and desktop — and sign off each case yourself

Description and use

git-qa lets an AI run tests across Android, iPhone, web, and desktop, while a human signs off each case, providing evidence and accountability.

Problem

Automated testing tools often claim to run all tests and report passes, but when failures occur the AI’s verdict is not accepted; accountability requires human verification.

Outcome

git-qa enables AI to execute tests across devices while a human authorizes each verdict, ensuring accountability and evidence of verification.

Workflow

• 1. Prepare a TSV test sheet with steps and expected results.
• 2. Install git-qa and Node 22+.
• 3. Launch git-qa, select device/URL/app and test sheet.
• 4. Watch the live screen as the AI runs each step.
• 5. Press 'D' to pass or 'F' to fail when the expected text is found or not.
• 6. Review run.json and screenshots for evidence of each verdict.
• 7. Optionally run unattended with '--no-ui' and review failures later.

Evidence

git-qa is an open-source human-in-the-loop QA runner built around that fact.

Source

An AI agent works through your test sheet; you watch the live screen and sign off each case with one key.

Source

You can run a sheet unattended overnight (--no-ui) and review only the failures in the morning.

Source